Battery Life DOD SOC Temperature Charge Rate Study

DrkAngel

1 GW
Joined
Dec 15, 2010
Messages
5,300
Location
Upstate-Western-Southern Tier NY. USA
I just came across
a decently thorough study testing
  • battery cycle life under various
  • battery charge levels,
  • charge rates
  • and temperature conditions

Cycling Aging of Lithium Ion Batteries
Aging study of state-of-art cells
04/06/2015 Mat4Bat Summer School


We all know that charging-discharging at higher temperatures is damaging to cycle life ...
this study demonstrates that charging-discharging at cold temperatures is horrifically worse!

Tests @1C discharge and various charge C-Rates
file.php


We all know that SOC higher charged voltage is damaging to cycle life ...
this study demonstrates that DOD Depth of Discharge is similarly damaging!
10%-90% vs 0%-80%

Tests @1C discharge and various charge C-Rates
file.php


A 10%-90% cycle is overwhelmingly preferable to a 0%-80% cycle!
 

Attachments

  • Lithtest cold.jpg
    Lithtest cold.jpg
    90.2 KB · Views: 1,528
  • Lithtest DOD-SOC.jpg
    Lithtest DOD-SOC.jpg
    92.5 KB · Views: 1,528
Wow. So a battery heater would be important in colder climates.

I've driven my EV in sub 5*C temps a few times. Wonder if it has a battery heater.
 
It's important to note that the discharge at lower temperatures is not likely causing the cycle life degradation as much as charging is.

You can kind of think of it like this, the lower the temperature, the lower the effective C rate. This means that charging at 1C is more like charging at 5C at lower temperature (5 degrees Celsius).

It is critical that battery packs are charged slowly, or not at all when the temperature gets near 0 degrees C. However, the discharge at lower temperatures is less of an issue, despite the fact that you will get much lower capacity.

It is true that a battery heater will yield a greater capacity at these temperatures (below 5C) , even if powered from the cells itself.

Fun data, thanks for sharing it.
 
P.S. I want to see this data such that it is not normalized to the full charge capacity at each temp, aka absolutely capacity, to see where the crossover points are.
 
grindz145 said:
You can kind of think of it like this, the lower the temperature, the lower the effective C rate. This means that charging at 1C is more like charging at 5C at lower temperature (5 degrees Celsius).

Yes this is it. If we look into the design guide of A123 LFP 20Ah cell you can see that maximum continuous charging current derating is: 3C at 25°C, 1C at 10°C, 0,5C at 0°C and 0,1C at -20°C. Also it is recomended to charge to lower voltage when charging below 0°C or above 40°C.
 
I also wonder what is the reason that they measured significantly worse results in 20-100% range (80% DoD) than in 0-100% range (100% DoD)?
 
Pajda said:
I also wonder what is the reason that they measured significantly worse results in 20-100% range (80% DoD) than in 0-100% range (100% DoD)?
The 20-100% range (80% DoD) than in 0-100% range (100% DoD) match fairly precisely at 1C and 3C charge rate, circles and triangles.
The major divergence is the 2C 20-100%, squares, for which there is no comparison.
0-100% not tested at 2C, no red square line.
 
Just a question for you battery experts that I can't find an answer for. We all know that shallower cycles and charging batteries to a lower voltage will result in extended cycle life. But is it voltage while charging or actual state of charge which makes the difference?

You can charge a li-ion cell to a SOC of 90% in 2 different ways. You can have 4.1v as a termination voltage, as in the charger will start the CV phase at 4.1v and never pass this voltage. Or you can charge it with a 4.2V charger and manually terminate the charger at a SOC of 90% about halfway into the CV phase. The problem with this method is that the cell will already have been at 4.2V for a while.

For example my smartphone. For me to charge this to 4.1V only I would have to unplug it at around 70% state of charge. Already at 80% it will have reached 4.2V at the battery and stay there for the remaining charge cycle.

I can't seem to find any information on this as every study I've seen seems to use for example a 4.1V charge vs a 4.2V charge.
 
mighty82 said:
Just a question for you battery experts that I can't find an answer for. We all know that shallower cycles and charging batteries to a lower voltage will result in extended cycle life. But is it voltage while charging or actual state of charge which makes the difference?

You can charge a li-ion cell to a SOC of 90% in 2 different ways. You can have 4.1v as a termination voltage, as in the charger will start the CV phase at 4.1v and never pass this voltage. Or you can charge it with a 4.2V charger and manually terminate the charger at a SOC of 90% about halfway into the CV phase. The problem with this method is that the cell will already have been at 4.2V for a while.

For example my smartphone. For me to charge this to 4.1V only I would have to unplug it at around 70% state of charge. Already at 80% it will have reached 4.2V at the battery and stay there for the remaining charge cycle.

I can't seem to find any information on this as every study I've seen seems to use for example a 4.1V charge vs a 4.2V charge.
I similarly would extol the benefits of charging at 4.20V and terminating as internal voltage attains 4.10V.
My advice would be a cutoff as input reaches a specific low current (Amps) level, rather than timed termination, due to varied DOD.
Toyed with idea of spring tensioned electromagnet powered by charge current. Just found possible circuit of promising viability, ordered one up for trial.
See - Plausible to build custom variable voltage Lion charger?
Reasonably, I would expect Internal high voltage as the major degradation factor with external high voltage as a minor or even null factor.
But, I have seen no actual study ...
At minimum, 4.2V charge voltage duration would be greatly reduced with an internal 4.10V termination, compared to extending charge till 4.20V.
 
As far as I can see, the board you mention will only take up to 30V. I would need at least 60V.

Is the internal voltage and external voltage not the same during charging? By the logic that external voltage doesn't matter, would charging 4.2V batteries with 4.3 or 4.4V to "skip" the cv phase not be the norm? Why do we have the cv phase at all?

I know LiFe batteries can be charged that way with a limited overvoltage until 100% soc and then termination.
 
DrkAngel said:
mighty82 said:
Just a question for you battery experts that I can't find an answer for. We all know that shallower cycles and charging batteries to a lower voltage will result in extended cycle life. But is it voltage while charging or actual state of charge which makes the difference?

You can charge a li-ion cell to a SOC of 90% in 2 different ways. You can have 4.1v as a termination voltage, as in the charger will start the CV phase at 4.1v and never pass this voltage. Or you can charge it with a 4.2V charger and manually terminate the charger at a SOC of 90% about halfway into the CV phase. The problem with this method is that the cell will already have been at 4.2V for a while.

For example my smartphone. For me to charge this to 4.1V only I would have to unplug it at around 70% state of charge. Already at 80% it will have reached 4.2V at the battery and stay there for the remaining charge cycle.

I can't seem to find any information on this as every study I've seen seems to use for example a 4.1V charge vs a 4.2V charge.
I similarly would extol the benefits of charging at 4.20V and terminating as internal voltage attains 4.10V.
My advice would be a cutoff as input reaches a specific low current (Amps) level, rather than timed termination, due to varied DOD.
Toyed with idea of spring tensioned electromagnet powered by charge current. Just found possible circuit of promising viability, ordered one up for trial.
See - Plausible to build custom variable voltage Lion charger?
Reasonably, I would expect Internal high voltage as the major degradation factor with external high voltage as a minor or even null factor.
But, I have seen no actual study ...
At minimum, 4.2V charge voltage duration would be greatly reduced with an internal 4.10V termination, compared to extending charge till 4.20V.

sorry i'm late to the party. i only recently started to seriously contemplating a motorcycle build. I study batteries, and perhaps i can shed some light on this matter..

the vast majority of capacity loss is due to parasitic side reactions of the electrolyte. the byproducts form a layer on the active material called the solid electrolyte interface. one of the reactants that form this byproduct is lithium, so one of the mechanisms of cycle life loss is the loss of lithium available for cycling. electrolyte decomposition is accelerated at higher voltages, so it is actually better to keep the batteries at around 40% SOC for greater longevity, and minimize the amount of time the battery spends at 100% SOC.

another failure mechanism is in the cathode material, which is a crystal structure. as the battery is charged, the lithium leaves the cathode and goes to the anode. if too much lithium leaves the cathode, the crystral structure can break down irreversibly, and lead to capacity loss.

on the other hand, if a battery is fully discharged, the anode can run into some problems. one problem is the ionic conductivity of the material decreases, so it can't accept charge quite as readily, especially in colder conditions. what happens when you try to charge the battery too fast when the lithium ions can't move that fast in the anode is you get lithium plating on the anode surface, which is BAD, and it will also decrease cycle life.

--------------------------------

that said, I've pondered about temperature management of the battery cells and the motor... It almost seems like an afterthought in most of the posts that I've seen. the nissan leaf modules appear to be a popular option. But the nissan leaf cars have had several battery life problems, where the battery degradation was faster than expected, due in part to poor temperature management. So I'm curious for those who have had diy bikes for a longer period of time, how has your battery life changed, and what did you do to manage it?

do people think about some sort of air cooling when they're building the pack? I see a lot of people put their cells inside metal boxes, but I feel like that's a terrible idea. I think it's much better to protect the batteries with grated material, to allow for better air flow during riding. even better, it would be better to make an air duct that will funnel air specifically to cool the batteries, for the proper heat exchanger effect. anyone have experience with that?

also, what about temperature management of the motor? I'm thinking about an AC 20, 72V 550Amp, on a 20s4p 34Ah pack (that's 4x34=136 Ah total, I'm still not quite sure how Ah is supposed to be expressed). It's supposed to take me up to 70 mph, and has more initial torque than my sv650, and at max amps, is only discharging at 2C. but i don't know how hot the motor would get.

I live in texas, so i don't really have to worry too much about cold weather, but it does get to freezing every once in a while. (also no garage :().


-------

I just realized I asked wayy too many questions in that one post, so I'll go ahead and start my own thread somewhere else. haha
 
Back
Top